Skip to content

Singapore · résumé

Ong Jun Xiong

AI Infrastructure & Backend Engineer

Download PDF

I build production agent runtimes, tool servers, and backend platforms. At TikTok, my work spans the shared AI layer and the Go services beneath global e-commerce operations.

junxiong.devEmailGitHubLinkedIn

Roles

Experience

TikTok

Singapore

AI Infrastructure Engineer, Promotion & Operations

Oct 2025 – present

  • Built the team's shared AI platform from zero: a LangGraph agent service on Google's A2A protocol, a Go MCP server exposing 16 analysis tools over the data warehouse, and a React SDK that adds streaming AI chat to a page in about five lines.
  • Adding LLM analysis to a new surface went from a multi-file change to a single config directory, and analysis latency dropped from minutes to seconds. In daily use by operators and analysts across the team.
  • Built the Marketing Strategy Center with operations teams across the US, Southeast Asia, and the UK: a Go/gRPC service and React console that enforces product eligibility rules and budget stop-loss continuously, replacing a daily manual review.
  • Shipped the Promotion Data Center analytics backend in Go with Kafka, Redis and Hive-backed OLAP: 26 statistics across 8 drill-down dimensions on 4 endpoints, serving ~13 markets with real-time alerting.

Software Engineer, Seller Center & Creator Marketplace

Apr 2024 – Oct 2025

  • Refactored hardcoded region logic so EU and future markets launch by configuration, and delivered the related compliance, data-pipeline and API migrations.
  • Built the Video Diagnosis 2.0 recommendation backend in Go (candidate generation plus personalised ranking for creator-product matching), validated through A/B tests before rollout.
  • Supported five owned and two contributed Go/gRPC services on the seller platform, handled 20+ on-call escalations, and added cache observability and alerting that stopped a recurring class of outages.

Compoze

Singapore

Founder

Jan 2025 – Dec 2025

  • Built and ran a one-person AI solutions company alongside TikTok. I handled discovery, scoping, pricing, delivery, and client training.
  • Shipped retrieval-augmented chat products with LlamaIndex, FastAPI, Next.js, and LlamaCloud. A client acquired the company at the end of 2025.

Earlier experience

Binance

Remote

Software Engineering Intern, Payments & Cards

Aug 2023 – Nov 2023

  • Shipped TypeScript and React UI, including the Crypto Box group-payment checkout flow with idempotent, double-confirm submission.

Thought Machine

Singapore

Forward Deployment Engineer (Intern)

May 2023 – Aug 2023

  • Forward-deployed on core-banking (Vault) client implementations. Built a Python remediation tool over the platform REST API with dry-run preview and idempotent apply, removing 20+ hrs/week of repetitive configuration work.

DBS Bank · Tanso

Singapore · Munich

Backend and full-stack engineering internships

2022

  • Built Java and Spring Boot services for 1,000+ daily B2B users at DBS, then a Next.js, tRPC and PostgreSQL carbon-accounting dashboard at Tanso that cut page load from 5s to under 50ms.

Tools

Skills

AI systems

MCP · A2A · LangGraph · LlamaIndex · RAG · agentic workflows

Backend

Go · TypeScript · Python · Java · gRPC · REST · Kafka · Redis

Data & infra

MySQL · PostgreSQL · ClickHouse · Hive · Docker · Kubernetes · AWS

Product

React · Next.js · distributed systems · client delivery · technical scoping

School

Education

National University of Singapore

Bachelor of Computing (Honours), Computer Science. Specialised in AI and software engineering; NUS Overseas Colleges in Munich.

2020 – 2023

Ong Jun Xiong

ENGINEER · BUILDER · SINGAPORE

EmailHobbiesArchiveNotesGitHubLinkedInSource

© 2026 Ong Jun Xiong